What’s New

These are new features and improvements notes for each release.

v0.9.0 (June 13, 2020)

  • Drop support / compatibility for Python 2.7.

  • inputs / outputs names are now optional.:

    @graph.register()
def f_my_function(a, b):
    c = a + b
    return c

  • Additional argument to Graph to prevent inputs deepcopy (@Tosa95). This is useful when having large inputs when the deepcopy might be slow, but remove inputs immutability.

v0.8.1 (October 27, 2019)

v0.8.0 (March 17, 2019)

  • Can pass list of Input / Output to Graph.

  • Names mapping.

  • Inputs validation with JSON Schema.

  • Some misc. refactoring and styling.

v0.7.0 (February 10, 2019)

  • Refactoring, and introduction of Input and Output objects.

  • Contracts feature.:

    graph.add_node(
    my_function,
    inputs=[Input(name='a', contract='>0'), Input(name='b', contract='float')],
    outputs=[Output(name='g', contract='float')]
)

  • Nodes IDs refactoring (now uses uuid).

  • Run topological sort only when needed.

  • Everything accessible from pyungo __init__ (@nelsontodd)

  • Docstrings / Sphinx doc / GitHub page

v0.6.0 (July 6, 2018)

  • Fix “false parallelism” implementation. This require an optional dependency (multiprocess module) which is more friendly to work with stateful objects and multiprocessing.

  • Ability to pass predefined inputs at the node definition:

    @graph.register(inputs=['a', {'b': 2}], outputs=['c'])
def f_my_function(a, b):
    return a + b

v0.5.0 (April 3, 2018)

  • Nodes can be run in parallel using Python multiprocessing module:

    graph = Graph(parallel=True, pool_size=3)

  • inputs and outputs parameters are mandatory and will raise an explicit error when missing.

  • Add a longer test using pvlib

v0.4.0 (March 23, 2018)

  • Fix single output when dealing with iterables. Previously, returning one output with an iterable was resulting in taking the first element.

  • Args and kwargs can be used in functions:

    @graph.register(
        inputs=['a', 'b'],
        args=['c'],
        kwargs=['d'],
        outputs=['e']
    )
def f_my_function(a, b, *args, **kwargs):
    return a + b + args[0] + kwargs['d']

  • Can create a Node without using the decorator:

    graph.add_node(f_my_function, inputs=['a', 'b'], outputs=['c'])

  • DAG pretty print (graph.dag)

v0.3.0 (March 18, 2018)

First version in pypi